This study will investigate the consequences of self-fertilization for the evolution of plant populations by either natural or artificial selection. The objectives are to (l) develop a mathematical model to predict response to selection in a partially selfing population as a function of genetic parameters; (2) analyze this model to determine under what genetic assumptions self-fertilization is expected to accelerate or retard selection; and (3) test the empirical adequacy of this model with a novel type of artificial selection experiment on the plant species Mimulus guttatus. The development of this model will aid field biologists in investigations of natural selection and should also facilitate the development of efficient techniques for the selective improvement of agricultural plants. Efficient plant breeding programs are critical to the maintenance and improvement of human health and nutrition standards.